In a rapidly evolving technological landscape, the rise of DeepSeek, a Chinese startup, is both fascinating and worrying. By launching R1, an open-source reasoning model, DeepSeek is directly taking on artificial intelligence giants, notably OpenAI. This development has resulted in meteoric adoption, exceeding expectations on platforms such as the App Store, and positioning DeepSeek as a major player in the innovation race. This phenomenon challenges the established hierarchy and invites us to rethink the future of artificial intelligence.<\/p>\n\n
Founded in May 2023 by Liang Wenfeng, the startup DeepSeek quickly made a splash in the world of AI. Just a few months after its creation, it launched the R1 model, which, according to numerous tests, outperforms OpenAI’s o1 model in key areas such as math and coding. But what does this innovation really represent? How did it trigger a wave of investor panic?<\/p>\n\n
DeepSeek has succeeded in producing a model that not only competes with existing solutions, but also offers a significantly lower cost of use. Access to the R1 API is set at $0.14 for one million tokens, compared to the $7.50 charged by OpenAI for similar services. This new pricing standard paves the way for an environment where artificial intelligence is more accessible to small businesses and researchers who lack the financial resources of organizations like OpenAI.<\/p>\n\n
To assess the capabilities of R1 compared to o1, consider the performance of each model on several benchmarks: <\/p>\n\n This difference in price and performance could transform the AI \u200b\u200blandscape, making advanced technologies accessible to a wider audience.<\/p>\n\n However, this meteoric rise is not without consequences. DeepSeek had barely begun to reap the benefits of its success when the startup had to limit new registrations due to a cyberattack. This type of risk underscores the fragility of this competitive field where data security is paramount. Furthermore, concerns about censorship persist, with Chinese models often having restrictions on certain sensitive topics.<\/p>\n\n It is essential to weigh these concerns against the advantages presented by DeepSeek. On the one hand, users can potentially benefit from an open-source model, and therefore transparency into how AI works; on the other, they must navigate an environment where government censorship is pervasive. The Geopolitical Implications of DeepSeek’s Rise<\/p>\n\n To date, several major technology companies are concerned about the potential implications of using Chinese AI tools. The following table summarizes the companies that have taken preventive measures:<\/p>\n\n Company <\/p>\n\n Toward a Future with DeepSeek and Open Source<\/p>\n\n The Benefits of Open Source in AI Development<\/p>\n\n Accessibility: <\/p>\n\n <\/p>\n\n <\/p>\n\n\n","protected":false},"excerpt":{"rendered":" In a rapidly evolving technological landscape, the rise of DeepSeek, a Chinese startup, is both fascinating and worrying. By launching R1, an open-source reasoning model, DeepSeek is directly taking on artificial intelligence giants, notably OpenAI. This development has resulted in meteoric adoption, exceeding expectations on platforms such as the App Store, and positioning DeepSeek as […]<\/p>\n","protected":false},"author":1,"featured_media":26821,"comment_status":"closed","ping_status":"","sticky":false,"template":"","format":"standard","meta":{"footnotes":""},"categories":[1398],"tags":[149,47070,1401,155,339],"class_list":["post-26883","post","type-post","status-publish","format-standard","has-post-thumbnail","hentry","category-news-ai-en","tag-artificial-intelligence-en","tag-chinese-startup-en","tag-deepseek-en","tag-openai-en","tag-technological-innovation-en"],"_links":{"self":[{"href":"https:\/\/www.mon-agent-ia.fr\/blog\/wp-json\/wp\/v2\/posts\/26883","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/www.mon-agent-ia.fr\/blog\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/www.mon-agent-ia.fr\/blog\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/www.mon-agent-ia.fr\/blog\/wp-json\/wp\/v2\/users\/1"}],"replies":[{"embeddable":true,"href":"https:\/\/www.mon-agent-ia.fr\/blog\/wp-json\/wp\/v2\/comments?post=26883"}],"version-history":[{"count":1,"href":"https:\/\/www.mon-agent-ia.fr\/blog\/wp-json\/wp\/v2\/posts\/26883\/revisions"}],"predecessor-version":[{"id":26884,"href":"https:\/\/www.mon-agent-ia.fr\/blog\/wp-json\/wp\/v2\/posts\/26883\/revisions\/26884"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/www.mon-agent-ia.fr\/blog\/wp-json\/wp\/v2\/media\/26821"}],"wp:attachment":[{"href":"https:\/\/www.mon-agent-ia.fr\/blog\/wp-json\/wp\/v2\/media?parent=26883"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/www.mon-agent-ia.fr\/blog\/wp-json\/wp\/v2\/categories?post=26883"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/www.mon-agent-ia.fr\/blog\/wp-json\/wp\/v2\/tags?post=26883"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}\n\n
\n \nCriteria<\/th>\n DeepSeek R1<\/th>\n OpenAI o1<\/th>\n<\/tr>\n<\/thead>\n \n Mathematics<\/td>\n Superior<\/td>\n Good<\/td>\n<\/tr>\n \n Coding<\/td>\n Having comparable results<\/td>\n Superior<\/td>\n<\/tr>\n \n Reasoning<\/td>\n Superior<\/td>\n Good<\/td>\n<\/tr>\n<\/tbody>\n<\/table><\/figure>\n\n Concerns about cyberattacks and censorship<\/h3>\n\n
DeepSeek’s rise takes place within a complex geopolitical context. The technological rivalry between China and the United States is exacerbated by recent tensions surrounding the regulation of technologies like TikTok and the impact this has on trade relations. Many companies, including Microsoft, have already imposed restrictions on the use of tools developed by DeepSeek.<\/h2>\n\n
\n\n
\n \nMeasure<\/th>\n Microsoft<\/th>\n<\/tr>\n<\/thead>\n \n Ban on Use<\/td>\n Google<\/td>\n<\/tr>\n \n Risk Assessment<\/td>\n Apple<\/td>\n<\/tr>\n \n Internal Ban<\/td>\n A Turning Point for Technological Innovation<\/td>\n<\/tr>\n<\/tbody>\n<\/table><\/figure>\n\n As AI development continues to accelerate, the question arises whether startups like DeepSeek could truly redefine the traditional balance of power in the industry. The availability of open source models could foster the emergence of new, more competitive solutions, surpassing established offerings.<\/h3>\n\n
With an open source reasoning model, DeepSeek embodies an approach that could revolutionize access to AI technology. By allowing researchers and developers to improve their own models using R1, the AI \u200b\u200becosystem could become more diverse and dynamic.<\/h2>\n\n
By offering the ability to customize AI models, DeepSeek not only fosters innovation but also greater collaboration within the developer community. Here are some key benefits of this approach:<\/h3>\n\n
DeepSeek’s privacy policy, which mentions that information may be stored on servers in China, highlights the need for international regulation of technology and data. Companies using R1 should be aware of these challenges to avoid potential impacts on user privacy.<\/h3>\n\n